Kolkata: Electoral outcomes in fiercely contested regions are often decided by razor-thin margins, where every vote counts. In Tamil Nadu, one seat was decided by one vote this time and in Rajasthan and Madhya Pradesh, even in the parliamentary elections, MPs have been elected with the margin of just one vote. Every vote matters.In West Bengal, the scale of the ‘special intensive revision’ or SIR was staggering. A massive 91 lakh names were struck from the voter list. This included 58 lakh routine deletions under the ASDD (absent, shifted, deleted, and displaced) categories during the draft revision, alongside another 27 lakh voters rendered ineligible following a judicial review of Under Adjudication (UA) cases. When paired with the 1.88 lakh new voter additions in the final roll published on February 28, 2026, the data points to a mathematically decisive role the revision exercise played. This exercise becomes important as during the SIR hearings before the apex court, Supreme Court judge, Justice Joymalya Bagchi had said, “If 10% of the electorate does not vote and the winning margin is more than 10%…what will happen? Suppose margin is 2% and 15% of [the] electorate who are mapped could not vote, then maybe…we are not expressing any opinion, but we would definitely have to apply our minds.”The constituency-level data suggests that the SIR was not just a routine roll-cleaning exercise in West Bengal. Its political significance comes from one central fact – in a large number of seats, the number of deleted voters was larger than the winning margin. This does not prove that every such result changed because of SIR, but it shows that roll revision directly entered the zone of electoral competitiveness.Deletions larger than winning margins in 150 seatsThe strongest indicator of the SIR having been decisive in dictating the political outcome is in seats where the winning margin was lesser than the total deletions, i.e. the combined ASDD and UA removals. When I combined routine ASDD deletions and the number of UA voters, the total number of deleted voters exceeded the final victory margin in a staggering 150 constituencies. West Bengal has 294 seats in all, which puts the number of such seats over the halfway mark in the assembly.Among these seats, the Bharatiya Janata Party can be seen to have had a clear advantage and led in 100, while the Trinamool Congress led in 48 and the Congress in two.In 2021, TMC had won 131 of these seats, and BJP, only 19. The two districts bordering Kolkata bore the absolute brunt of the roll contraction, accounting for nearly 30% of all highly affected seats. In the North 24 Parganas, the TMC dominated in 2021 by winning 23 of the 26 affected seats, but the landscape completely flipped in 2026 as BJP captured 21 of them. Similarly, in the South 24 Parganas, the TMC had previously swept all 19 of these seats, but the BJP made deep inroads post-SIR, flipping 10 of them.Beyond these epicentres, the net roll contraction heavily impacted Muslim-rich and highly competitive districts. In Murshidabad, the TMC’s 2021 tally of 13 out of 15 affected seats fell to just 6, with the BJP picking up 7 and the Congress claiming 2. In Purba Bardhaman, the TMC lost 11 of its 13 previously held seats to the BJP. This trend continued in Howrah and Hooghly; across their 22 combined affected seats, all of which were won by the TMC in 2021, the BJP managed to capture 14 in the recent elections.Urban centres were not immune to this sweeping pattern either. In Paschim Bardhaman, the BJP swept all 8 highly contracted seats, 5 of which were previously held by the TMC. Even in the capital districts of Kolkata North and South, the BJP managed to wrest 6 of the 11 affected seats away from the TMC’s 2021 clean sweep including Bhabanipur, where outgoing Chief Minister Mamata Banerjee lost to Suvendu Adhikari. Ultimately, the geographical distribution clearly indicates that the roll contractions were not randomly scattered. They overwhelmingly hit the TMC’s strongest fortresses from the previous election, enabling the BJP to successfully flip 100 of these 150 highly contracted battlegrounds.In Satgachhia, the victory margin was a mere 401 votes for the BJP candidate. However, the total deletions were massive, with 17,669 ASDD deletions and 8,785 UA voters found ineligible – totalling over 26,000 deletions. Similarly, in Rajarhat New Town, a margin of 316 votes was dwarfed by over 63,000 total deletions. In Raina, BJP led by 834 votes, while total deletions crossed 23,000.Did the areas with the highest ASDD deletions favour BJP?If only ASDD deletions are considered, and not UA, the impact is still enormous. In 110 constituencies, ASDD deletions alone exceeded the winning margin. In these seats too, the BJP benefited disproportionately, winning 72 seats, exactly double the 36 seats won by the TMC. The Congress won 2 seats. The geographic concentration mirrors the overall deletions, with North 24 Parganas (19 seats) and South 24 Parganas (17 seats) topping the list. In 2021, the TMC won 102 of them compared to just 8 for the BJP. To isolate the direct change, the data shows that 66 of these 110 seats actually changed hands between the two elections. The TMC lost every single one of these 66 flipped constituencies, with the BJP capturing 64 of them and the Congress gaining the remaining two. Ultimately, the data demonstrates that routine ASDD deletions were undeniably concentrated in seats the TMC had previously won, and in areas where these roll cleanups outnumbered the winning margin, the TMC’s prior dominance was effectively shattered. Together, the twin districts of North and South 24 Parganas accounted for nearly a third of all seats heavily impacted by ASDD deletions. In North 24 Parganas, the TMC had nearly swept in 2021 by winning 18 out of 19 seats, but the landscape flipped drastically in 2026 as BJP captured 15 of them, reducing the TMC to just four. Similarly, in South 24 Parganas, the TMC’s perfect 16-0 sweep from 2021 was broken when the BJP made significant inroads post-revision to capture eight of these tightly contested areas. In Howrah and Hooghly, the TMC had won all 12 affected seats and all eight affected seats, respectively, during the previous election. By 2026, the BJP won 6 of such seats in Howrah and almost overrun Hooghly by capturing 7. In the Muslim-heavy district of Murshidabad, the TMC saw its 2021 tally of 9 out of 10 highly affected seats plummet to just 2, with the BJP capturing 6 and the Congress taking the remaining 2. Urban clusters were also deeply impacted. In Paschim Bardhaman, the BJP swept all 6 affected seats, 5 of which were previously held by the TMC, while in Kolkata North, the BJP wrested four seats away from the TMC’s prior clean sweep. UA deletions in Muslim-heavy seats helped BJPThe exclusion of voters marked ‘UA’ shows a more politically sensitive pattern. In 49 constituencies, as already reported by The Wire, the number of UA deletions exceeded the winning margin. Unlike the broader ASDD category, the political impact here was much more evenly split. The BJP won 26 of these seats, while the TMC captured 21, and the Congress took 2. However, when compared to 2021 results, the impact looks severe.These constituencies had an average Muslim population of 33.69%, much higher than in the other deletion categories. Murshidabad was the most affected district, with 8 seats where voters marked UA exceeded the winning margin. Other affected districts included North 24 Parganas, South 24 Parganas, Purba Bardhaman and Hooghly.Historically, TMC dominated nearly 100% of these high-minority-concentration areas, having won 48 of the 49 seats in 2021, with BJP holding just one.However, the 2026 results completely transformed this landscape. BJP emerged victorious in 26 of these seats (a net gain of 25), while TMC’s hold was reduced to just 21 seats, suffering a steep net loss of 27 seats overall. Additionally, Congress managed to pick up 2 seats. Ultimately, the data indicates that the deletion of under-adjudicated voters, intersecting with exceptionally tight winning margins, was a pivotal factor that resulted in the TMC losing more than half of these stronghold constituencies to the BJP. The examples are striking. In Samserganj, UA deletions were 74,775, far above TMC’s margin of 7,587. In Raninagar, Congress won by 2,701 votes, while UA deletions were 17,140. In Jangipur, BJP won by 10,542, but UA deletions stood at 36,581 in a constituency with over 51% Muslim population. In Purba Bardhaman’s Raina, the BJP secured a victory margin of 834 votes, but the number of UA voters found ineligible was a whopping 11,284, showing a massive gap between the victory spread and the targeted deletions. Had the UA deletions not taken place, five seats would have flipped – four from BJP to TMC and one, from Congress to TMC.This is not to say that nothing has changed politically since 2021, or that all those deleted would have voted in one direction, but when there are deletions at such a large scale and winners are decided by a razor’s edge, who SIR has pushed out has to be looked at closely.Fresh enrolments have marginally favoured BJP While deletions dominated the narrative of tight races, the addition of new voters also played a decisive, albeit highly concentrated, role. In just five constituencies, Satgachia, Rajarhat New Town, Indus, Raina and Mandirbazar, the number of newly added voters in the final roll exceeded the margin of victory, meaning these fresh enrolments hypothetically held the power to sway the final result. The BJP was the primary beneficiary here, winning 4 of the 5 seats, while the TMC won 1. These seats have a high average Scheduled Caste population of 31.14%, suggesting that additions mattered most in a narrow cluster of SC-heavy or semi-rural constituencies.In South 24 Parganas’ Satgachia, the 401-vote margin was easily eclipsed by the 3,023 total additions to the electoral roll. In 2021, the TMC won the seat with a margin of 23,318 votes. Similarly, in Rajarhat New Town, the 316-vote margin was surpassed by 2,543 total additions. In 2021, the TMC’s margin in this seat was a staggering 56,432.What if…To understand the true weight and consequence of the SIR, we must pose a critical hypothetical question: What would the 2026 electoral landscape look like if these sweeping electoral roll revisions had never occurred? It is impossible to know exactly which candidate any specific added or deleted individual voted for.Therefore, this simulation acts as a mathematical stress test. It calculates the maximum net SIR impact (total additions + ASDD deletions + UA deletions) and compares it directly against the final winning margins. First caseIn the first case, the model assumes a worst-case scenario for the winning candidate, that every deleted voter would have voted for the runner-up, and every newly added voter cast their ballot for the winner.If we isolate the 137 seats that actually changed hands between the 2021 and 2026 elections, the data reveals that 87 of those flips could hypothetically be attributed directly to the roll revisions.If the SIR had not happened, and operating under the worst-case electoral scenario described above, these 87 seats could have theoretically remained with the party that won them five years prior.All 87 of these highly contested, mathematically vulnerable seats were won by the TMC in 2021. But in 2026, BJP captured 84 of these vulnerable seats, while Congress captured 2.Without the SIR, the mathematical possibility exists that the TMC could have successfully defended these 87 constituencies. If those seats had held firm, the ultimate composition of the assembly would look drastically different, significantly shrinking the BJP’s total gains.Simulation examples: Overwhelming SIR footprintsTo grasp how deeply the roll modifications penetrated these races, we can look at specific constituencies where the TMC incumbent was unseated by the BJP. In many of these cases, the sheer volume of additions and deletions did not just edge past the victory margin, but eclipsed it by massive ratios.In seats like Hemtabad and Kushmandi, the deletions were roughly three times the size of the margin. The most striking example is Karandighi. While a victory margin of nearly 20,000 votes would traditionally signal a definitive mandate, the electoral roll there saw an overwhelming deletion of 52,807 voters!While this model is an analytical tool measuring vulnerability rather than real-world voting, it points towards a reality: The electoral roll revisions were not just procedural maintenance, they were undeniably large enough to have rewritten the final outcome in the state’s tightest battlegrounds.Second caseIn the second case, a ‘no-SIR counterfactual model’ was used. The model does not assume that all deleted voters would have voted against the eventual winner. Instead, it uses each constituency’s 2021 assembly election voting pattern as the baseline.The model follows three steps. First, all deleted voters, UA and ASDD, are added back to the electorate. Second, these restored voters are distributed among parties according to the constituency’s 2021 party-wise vote share, meaning, the deleted voters are assumed to follow the 2021 voting pattern.Third, SIR-era additions are removed using the 2026 party distribution, so that the simulation isolates the effect of roll revision as far as possible.This is not a claim of exact voter behaviour. It is a constituency-level stress test. Were the deletions large enough, and politically distributed enough, to plausibly alter outcomes?Under this hypothetical ‘no-SIR’ model, 11 seats flip, all to TMC. BJP’s tally falls from 207 to 198, TMC rises from 80 to 91, and Congress falls from 2 to 0 in the 293 valid constituencies analysed. Falta was excluded as it is yet to be counted.Of the 11 flipped seats, 9 seats move from BJP to TMC and 2 from Congress to TMC. The flipped constituencies include Farakka, Jangipur, Raninagar, Rajarhat New Town, Satgachhia, Tollyganj, Jorasanko, Kashipur-Belgachhia, Champdani, Jangipara and Raina.Some examples from this hypothetical exercise are striking. In Rajarhat New Town, a BJP lead of 316 becomes a TMC lead of nearly 14,959. In Tollyganj, BJP’s margin of 6,013 becomes a TMC lead of 3,603. In Jangipur, BJP’s lead of 10,542 turns into a TMC lead of 12,003.ConclusionThe SIR did not overturn the overall result; BJP would still remain ahead in our simulation of ‘no SIR’. But it likely mattered in a decisive cluster of close contests. The data suggests that deletions, especially ASDD deletions, disproportionately helped BJP in margin-sensitive seats, while UA deletions had a sharper impact in minority-heavy constituencies. In a state where many constituencies are decided by narrow margins, SIR, just its statistical impact, definitely became an electoral factor in itself.